May 10,1963
@ Copyrighl 196s by the American Chemical Society
The Use of Amines i n Liquid-Liquid Extractions of Nucleic Acids and Related Compounds JOSEPH X. KHyaa From the Biology Division, Oak Ridge National Laboratory,* Oak Ridge, Tennessee Received November 9,1962
Liquid anion exchange procedures were developed and tested as a method of investigating nucleic acids and lower-weight components that are associated with them. The liquid exchangers are high-molecular-weight amine salts dissolved in water-immiscible solvents. Upon contact with an aqueous salt solution, the anions in the organic phase can be replaced by those contained in the aqueous phase. Di-, tri-, and tetranucleoside phosphates, RNA, DNA, and lower-weight oligonucleotides are extracted readily from dilute acetate or formate buffer solutions and are recovered from the organic phases by increasing the aqueous buffer concentration or by using dilute alkaline solutions that neutralize the amines and allow extracted compounds to reappear in the aqueous phase. Inherent differences in the extraction coefficients for a given series of compounds indicate that this technique can be used to make practical separations. Paraffin chain quaternary ammonium salts were
first used in nucleic acid chemistry by Jones (1953), who prepared nucleic acids v i a an insoluble c2tyltrimethylammonium complex. Aubel-Sadron et al. (1960, 1961) and, from the same laboratory, Weil and Ebel (1962) have systematically studied the action of several types of quaternary ammonium salts and have shown a correlation between the molecular weight of a given series of quaternary amines and their ability to precipitate nucleic acids from aqueous solutions. I n addition these workers showed that the precipitated quaternary ammonium salts of nucleic acid are soluble in several organic solvents. Their results indicate that a separation of different types of RNA may be possible because of differences in the solubilities of quaternary RNA complexes in a given organic solvent. Liquid ion exchange was first used by Smith and Page (1948), who used long-chain high-molecular-weight amines as acid-binding extractants. This technique has been used with remarkable success by Coleman et al. (1958) for developing methods for recovering uranium and related metals from ore leach liquors. This technique is similar in approach to, but differant in principle from, the countercurrent distribution system initiated by Warner and Vaimberg (1958) for the partition of RNA and also polynucleotides between two liquid phases. The partition technique was further developed by Holley et al. (1960) and also Kirby (1960) and refined for the fractionation of yeast amino acid acceptor RNA by Doctor et al. (1961). In the liquid anion exchange method, amines dissolved in water-immiscible solvents are converted to a desired salt form either by direct treatment with concentrated acid or by shaking with aqueous solutions of a given acid or buffer solution. Thus converted, the anion of the amine salt can be replaced by another
anion by subsequent equilibration with a fresh aqueous phase. As pointed out by Kunin (1961),liquid ion exchange differs from solvent extraction in that liquid ion exchange depends upon a reaction or energy of binding between the solute and the extractant. The over-all reaction is analogous to those of conventional solid anion exchangers. Liquid anion exchange studies have been initiated t o test the feasibility of this technique for (1)the extraction of RNA, DNA, nucleoside mono-, di-, and triphosphates, and intermediate-molecular-weight oligonucleotides from aqueous salt solutions, and (2) the separation of the compounds listed from each other as well as further separations within a given group.
EXPERIMENTAL PROCEDURES General Considerations.---In this paper, only singlebatch distribution studies are reported. The compounds used were dissolved in aqueous salt solutions and shaken with equal volumes of organic phases containing water-immiscible amine salts. Most of the extractions were carried out in small separatory funnels equipped with Teflon drainage plugs. The phases were agitated with a wrist-action shaking machine. For smaller volumes (3% in the preparation gives a gel-like aqueous phase with several of the aminesolvent systems tested, but more than 80 % of the RNA can be extracted into the organic phase. For DNA preparations high in protein, it is difficult to distinguish any phase separation owing to the formation of a bulky precipitate that extends into both phases. Furthermore, when nucleic acid of low protein content is extracted, it is difficult to maintain two clear phases when the aqueous salt concentration is increased to obtain a series of decreasing extraction coefficients. A t a buffer concentration in the range of 0.8 to 1.0 M, precipitates are formed which layer at the interface and may contain as much as 15% of the nucleic acid. I n extractions of nucleoside triphosphates several primary amines can be used interchangeably without apparent changes in their extraction behavior; however, the same amines differ markedly in their ability to extract the nucleic acids. Table I11 demonstrates the 3%) and extraction behavior of RNA-I1 (protein DNA (protein 1%) with sweral primary amines. Extraction of Polyadenylic Nucleotides.-Preliminary experiments were carried out on some nucleotides given to us by Dr. Leon Heppel. Only small quantities were available, and the extraction of these compounds was carried out by mixing 1 ml of 0.1 M 3-ethyl-lisobutyloctyl formate in butyl ether with the compounds dissolved in 1 ml of 0.5 M formate buffer, p H 3.5. For comparison, adenosine tetraphosphate was also included in this series. These results are given in Table IV.
-
-
DISCUSSION Structure and type of amine are two variables that are being tested for any selectivity that amines may have for the extraction of a particular class of compounds. At present the only indication that structure may be related to selectivity is the different behavior of a series of primary amines in their reactions with RNA and DNA (Table 111). However, with a given amine-solvent system, inherent differences in extraction coefficients are readily obtained in the lower-
Phase % Observa- Extionsb tracted -
Amine.
Solvent
JM-T JM-T
Butyl ether Butyl ether Butyl ether Butyl ether Isoamyl acetate Isoamylacetate Isoamyl acetate Isoamyl acetate Isoamyl acetate Isoamyl acetate Isoamyl acetate Isoamyl acetate
1 1 1 1 2 2 3 3
4 4
T.F. T.F. T.F. T.F. Ppt. I Ppt. I Ppt. 0 Ppt. I Ppt. I Ppt. I Ppt. A ppt. I
99 97 92 97 54 90
--
0 25 0
79 50
1 = 3-ethyl-1-isobutyloctyl;2 = 3,5,5-trimethylhexyl; = 1-(3-ethylpentyl)-4-ethyloctyl; 4 = 1-nonyldecyl. T.F. = thin film a t interface; ppt. I = precipitate at interface; ppt. 0 = precipitate in organic phase; ppt. A =
3
*
precipitate in aqueous phase.
TABLE IV EXTRACTION OF POLYADENYLIC NUCLEOTIDES AND ADENOSINETETRAPHOSPHATE Organic phase: 0.1 M 3-ethyl-1-isobutyloctyl formate in butyl ether. Aqueous phase: 0.5 M formate buffer, pH 3.5. Mixing of phases by a Vortex Jr. mixer for 2 minutes, phase ratio 1 : l . % Compound
(PAh (PA)3 (PA)4 (PA)5 Adenosine tetrapod
Extracted 7 75 81
88 98.5
EA0
0.08 3 4 7 68
molecular-weight phosphate compounds, so that a given pair of compounds, e.g., ADP-ATP or PA)^(pA)s can be separated easily with a few batch extractions. For other pairs or groups of compounds with separation factors (ratio of EAo of one compound to another) of much lower magnitude, countercurrent distribution would be necessary to effect practical separations. It is also obvious from the data that the molecular weight is an important factor to consider in using an amine as an extractant. Amines of molecular weight as low as 185 can remove RNA and other compounds from dilute buffer solution, but this is about the lowest limit that is practical. Losses of amine salt to the aqueous phase and complete loss of extractant power occur when lower-weight amines are used. The mechanism of extraction by this technique is not completely understood (Coleman et al., 1958). However the analogy of conventional ion exchange involving sorptions by alkylammonium groups in solid anion exchanges is very striking (Kunin, 1961). For instance, different salt forms of an amine show particular affinities in their extractability for a given compound, and, as the charge in a particular class of compounds is increased, correspondingly higher extraction coefficients are obtained. In an extraction sequence, a series of decreasing extraction coefficients can be obtained when the aqueous salt concentration is increased. Also conversion of an amine salt to free amine im-
AMINES
Vol. 2,No. 3, May-June, 1963
AS LIQUIDANION
EXCHANGERS
405
CI
i
E
0
(D
cu
G
$220 \
v,
z 180
/
3
/O
v
5z 140
s 100 w
I-
a I
%
60
0 I
a
E a W
20 0
P 10
E U
50 70 90 110 130 150 170 190 210 AQUEOUS TRIPHOSPHATE CONCN. (UNITS/ml AT 260mp)
30
230
FIG.4.-Nucleoside triphosphate isotherms. Cascade extractions (organic phase contacting successive volof aqueous phase) a t phase ratio 1:1 and contact time of 10 minutes per stage by wrist-action shaker 0, ATP in 0.5 M formate buffer, p H 3.5, us. 0.1 Y 1-(3-ethylpenty1)-4-ethyloctylamine formate in butyl ether. A, UTP in 0.25 M formate buffer, pH 3.5, us. 0.1 M 3-ethyl-1-isobutyloctylamine formate in butyl ether. Arrows indicate stage at which aqueous phase concentration of nucleoside triphosphates equals initial concentration. Additional points after arrows are the concentrations of a different solution of triphosphates in which the aqueous phase far exceeded initial stock concentrations. umes
mediately releases extracted compounds and prevents the amine from extracting any additional material. All the foregoing observations fit qualitatively an aR-NHJF,,
+
Xsq-O
+
_J (R-NH~)axorg
~ F a q -
E a O = K(RNH3F). _ _ - ~ (F -)a
equation which can be derived for the liquid ion exchange reaction. Thus if F is equal to formate and X is equal to a substance such as UTP (where C, = concentration of UTP in a given phase), then a t a given formate concentration the EAo for UTP would be expected to be proportional to some power of the concentration of free amine salt used as the extractant, and at a given amine concentration the EnO should vary inversely with the formate concentration. At p H lower than 3.5, UTP would be expected to have a negative charge of at least 3, and therefore the extraction coefficient should vary as the cube of the free amine salt concentration. However, as seen in Figure 1, the extraction coefficient of UTP and of both RNA and ATP varies directly to the first power of the amine salt concentration. Also,as shown in Figure 4, a limiting association number of about seven moles of amine to one mole of triphosphate was determined from the extraction isotherms. This same limiting number was found for calf liver RNA saturated in the organic phase with the amine JM-T. Thus a simplified interpretation of liquid ion exchange does not adequately explain the empirical relationship of the data as found, but the reactions considered as liquid ion exchange
have been useful in designing and correlating the various extraction experiments. Liquid-liquid extractions involving the use of amines as extractants are ideally suited for countercurrent distribution or partition chromatography techniques. However some means of overcoming the occurrence of gel, or third phase, formation must be developed for the adoption of this technique in the separation of high-molecular-weight phosphate polymers especially when they are contaminated with protein. ACKNOWLEDGMENTS The author wishes to thank Dr. E. Volkin for his interest and cooperation during the course of these investigations and in particular acknowledge the suggestions and direction given me during the isolation of various nucleic acid preparations. Gratitude is expressed to Dr. C. F. Coleman, W. D. Arnold, F. D. Seeley, and D. J. Crouse of the Chemical Technology Division of Oak Ridge National Laboratory for their generous supply of several amines used in these studies. Also for the same reason to the following commercial establishments: General Mills, R o b and Haas, Universal Oil, Archer-Daniels-Midland, Foremost Chemical, and the Armour Chemical Companies. REFERENCES Astrachan, L., Volkin, E., and Jones, M. H. (1957), J. Am. Chem. Soc. 79, 130. Aubel-Sadron, G., Beck, G., and Ebel, J. P. (1961),Biochim. Biophys. Acta 53, 11. Aubel-Sadron, G., Beck, G., Ebel, J. P., and Sadron, C. (1960), Biochim. Biophys. Acta 42, 542. Brown, K. B., Coleman, C. F., Crouse, D. J., and Moore, J. G. (1954), AEC Report ORNL-1734, p. 24.
406
Biochemistry
SAM KATZ AND F. ELLINGER
Coleman, C. F., Brown, K. B., Moore, J. G., and Crouse, D. J. (1958), Ind. Eng. Chem. 50, 1756. Doctor, B. P., Apgar, J., and Holley, R. W. (1961),J. Biol. Chem. 236, 1117. Holley, R. W., Apgar, J., and Doctor, B. P. (1960), Ann. N. Y. Acad. Sei. 88, 745. Jones, A. S. (1953), Biochim. Biophys. Acta 10, 607. Kirby, K. S. (1960),Biochim. Biophys. Acta 40, 193. Kunin, R. (1961),Amber-Hi-Lites, No. 62, Rohm and Haas Company, Philadelphia, Pa. Lowry, 0. H., Rosebrough, N . J., Farr, A. L., and Randall, R. J. (1951),J. Biol. Chem. 19%3,265.
Sevag, M. G., Lackman, D. B., and Smolens, J. D. (1938), J. Biol. Chem. 124, 425. Scott, J. E. (1960), in Methods of Biochemical Analysis, vol. 8, Glick, D., editor, New York, Interscience Publishers, Inc., p. 156. Smith, L. E., and Page, J. E. (1948), J. Chem. Soc. Chem.
I d . 67, 48. Volkin, E., and Carter, C. E. (1951),J. A m . Chem. Soc. 73, 1516. Warner, R. C., and Vaimberg, P. (1958),Fed. P m . 17, 331. Weil, J. H., and Ebel, J. P. (1962), Biochim. Biophys. Acta 55, 836.
Preparation of Isoionic Protein by Electrodialysis with Permselective Membranes* SAM
KATZAND F.
ELLINGERt
From the Naval Medical Research InstLtute, National Naval Medical Center, Bethesda, Maryland Received November 26, 1962
A water-jacketed apparatus employing permselective membranes for the electrodialysis of water-soluble polymers is described. This apparatus, characterized by its simplicity of construction and operation, is singularly free from leaks because of the incorporation of O-rings in its design. Heat production, type and concentration of salts, electrolysis products, geometry, membrane type and resistance, stirring, etc., are evaluated with respect to the rate and efficiency of the electrodialysis process. Routinely, a 2% solution of albumin in 0.10 M NaCl can be deionized in 4 hours. Two voltage effects, namely, coulombic heat and membrane polarization, are discussed in relation to the denaturation of protein. The values for the b i o n i c points of some proteins were determined and sujbeded to a critical analysis. In spite of the oblrious advantage of preparing saltfree protein by electrodialysis this technique has failed to gain wide acceptance for szveral raasons: electrodialysis with cellophane membranes generally required from 10 to 30 hours for adequate desalting, and the high voltages employed often resulted in denatured product. Frequently the conventional apparatus was prone to leak, resulting in a material loss and the hazard associated with electrical malfunction. With the advent of commercially available permselective membranes the initial objections are no longer applicable. The instrument described is simple in design and operation, free from leaks, and can readily accommodate volumes ranging from 5 to 35 ml. An analysis is made of the electrodialysis process, with particular reference to the factors which serve to set an upper limit to the current and or voltage which can be applied to this system.
The capacity of the water-jacketed center cell is 35 ml. Its contents can be stirred by either magnetic or mechanical means. For processing smaller volumes, cylindrical glass inserts are placed in this chamber. Burets or electrodes can be introduced through the top ports for p H monitoring, while the bottom port serves as the outlet. The membranes used were Nepton-AR-111-A (anion exchange) and Nepton
APPARATUS A water-jacketed electrodialysis unit for use with permselective membranes consisting of two identical glass electrode vessels and a Plexiglas center or sample cell was designed (Fig. 1). Each electrode cell, with a capacity of 15 ml, has shiny platinum electrodes, 28 mm diameter, mounted 6 mm from the membrane surface. In operation, influent distilled water is directed laterally against tke membrane and electrode surfaces to prevent the accumulation of electrolysis products and to minimize temperature gradients.
* This paper was presented in part at the 140th Meeting of
the American Chemical Society at Chicago, Illinois, September, 1961 The opinions or assertions contained herein are the private ones of the authors and are not to bi. construed as official or reflecting the views of the Navy Department or the Naval service at large. t Deceased.
FIG. 1.-An expanded view of the electrodialysis unit. A , electrode cell; B , center or sample cell; I, polyethylene connector for water inlet; 2, platinum electrode; 3, needle valve for electrode water outlet; 4, polyethylent support for membrane; 5 , rubber gaskets; 6, permselectike membrane; 7, O-ring; 8, cooling water inlet and outlet; 9, electrode ports; 10, sample outlet; 11, water jacket. The remainder of the unit is identical to the labeled portion.
